4c9cc558 |
1 | #' Plot |
2 | #' |
3 | #' It is a function which plots relevant parameters |
4 | #' |
a6b60f91 |
5 | #' @param model the model constructed by valse procedure |
6 | #' @param n sample size |
4c9cc558 |
7 | #' @return several plots |
8 | #' |
9 | #' @examples TODO |
10 | #' |
11 | #' @export |
12 | #' |
a6b60f91 |
13 | plot_valse = function(model,n){ |
4c9cc558 |
14 | require("gridExtra") |
15 | require("ggplot2") |
16 | require("reshape2") |
a6b60f91 |
17 | require("cowplot") |
4c9cc558 |
18 | |
a6b60f91 |
19 | K = length(model$pi) |
4c9cc558 |
20 | ## regression matrices |
21 | gReg = list() |
22 | for (r in 1:K){ |
23 | Melt = melt(t((model$phi[,,r]))) |
24 | gReg[[r]] = ggplot(data = Melt, aes(x=Var1, y=Var2, fill=value)) + geom_tile() + |
25 | scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0, space = "Lab") + |
26 | ggtitle(paste("Regression matrices in cluster",r)) |
27 | } |
28 | print(gReg) |
29 | |
30 | ## Differences between two clusters |
31 | k1 = 1 |
32 | k2 = 2 |
33 | Melt = melt(t(model$phi[,,k1]-model$phi[,,k2])) |
34 | gDiff = ggplot(data = Melt, aes(x=Var1, y=Var2, fill=value)) + geom_tile() + |
35 | scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0, space = "Lab") + |
36 | ggtitle(paste("Difference between regression matrices in cluster",k1, "and", k2)) |
37 | print(gDiff) |
38 | |
39 | ### Covariance matrices |
40 | matCov = matrix(NA, nrow = dim(model$rho[,,1])[1], ncol = K) |
41 | for (r in 1:K){ |
42 | matCov[,r] = diag(model$rho[,,r]) |
43 | } |
44 | MeltCov = melt(matCov) |
45 | gCov = ggplot(data =MeltCov, aes(x=Var1, y=Var2, fill=value)) + geom_tile() + |
46 | scale_fill_gradient2(low = "blue", high = "red", mid = "white", midpoint = 0, space = "Lab") + |
47 | ggtitle("Covariance matrices") |
48 | print(gCov ) |
49 | |
50 | ### proportions |
51 | Gam = matrix(0, ncol = K, nrow = n) |
52 | gam = Gam |
53 | for (i in 1:n){ |
54 | for (r in 1:K){ |
55 | sqNorm2 = sum( (Y[i,]%*%model$rho[,,r]-X[i,]%*%model$phi[,,r])^2 ) |
56 | Gam[i,r] = model$pi[r] * exp(-0.5*sqNorm2)* det(model$rho[,,r]) |
57 | } |
58 | gam[i,] = Gam[i,] / sum(Gam[i,]) |
59 | } |
60 | affec = apply(gam, 1,which.max) |
61 | gam2 = matrix(NA, ncol = K, nrow = n) |
62 | for (i in 1:n){ |
63 | gam2[i, ] = c(gam[i, affec[i]], affec[i]) |
64 | } |
65 | bp <- ggplot(data.frame(gam2), aes(x=X2, y=X1, color=X2, group = X2)) + |
a6b60f91 |
66 | geom_boxplot() + theme(legend.position = "none")+ background_grid(major = "xy", minor = "none") |
67 | print(bp ) |
4c9cc558 |
68 | |
69 | ### Mean in each cluster |
70 | XY = cbind(X,Y) |
71 | XY_class= list() |
72 | meanPerClass= matrix(0, ncol = K, nrow = dim(XY)[2]) |
73 | for (r in 1:K){ |
74 | XY_class[[r]] = XY[affec == r, ] |
75 | meanPerClass[,r] = apply(XY_class[[r]], 2, mean) |
76 | } |
77 | data = data.frame(mean = as.vector(meanPerClass), cluster = as.character(rep(1:K, each = dim(XY)[2])), time = rep(1:dim(XY)[2],K)) |
78 | g = ggplot(data, aes(x=time, y = mean, group = cluster, color = cluster)) |
79 | print(g + geom_line(aes(linetype=cluster, color=cluster))+ geom_point(aes(color=cluster)) + ggtitle('Mean per cluster')) |
80 | |
81 | } |